[rrd-users] Trying to find out if rrd would be suitable for this task.

[David Thornton]

You add but never delete.
And actually you are not really adding , you are filling an already existing
bucket with fresher data.
Think of it as a table with a fixed number of rows that get updates
in sequence until you get to the last row, you then go back to the beginning
and start updating the first row again.
the a row has only two columns: time and value.

I whipped up in 15 minutes a perl script that takes a "SHOW GLOBAL
VARIABLES" from mysql and greps out variables I care about then rrdupdates a
single rrd.

Ihe question about "independant varaibles" really is about when you collect
the data. If you collect it all at the same time every time, then useing one
process and one rrd is the way to go.

If some variables should be recorded once an hour and some once every five
minutes say, then you should have an rrd for each frequency of collection (
and tune each rrd appropriately ).

David



On Fri, Sep 24, 2010 at 8:19 PM, Tim Uckun <timuckun at gmail.com> wrote:

> >
> > Is this period the same for each, or do they have to be independent?
>
> I think they would be independent. What I would do is to send data to
> rrd when each record is changed.
> >
> > Yes you can.  Multiple items can be tracked in a single database.  You
> > just add in additional data sources (DS) when you create the database.
>
>
> This is pretty dynamic? In other words is it like adding, or deleting
> rows from a table?
>
> _______________________________________________
> rrd-users mailing list
> rrd-users at lists.oetiker.ch
> https://lists.oetiker.ch/cgi-bin/listinfo/rrd-users
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.oetiker.ch/pipermail/rrd-users/attachments/20100924/aacae82d/attachment-0001.htm